What is Monte Carlo Tree Search? - Artificial Intelligence
64.6 هزار بار بازدید -
7 سال پیش
-
Monte Carlo Tree Search is
Monte Carlo Tree Search is a search technique in Artificial Intelligence. This has recently been used by Artificial Intelligence Programs like AlphaGo, to play against the world's top Go players. Monte Carlo methods have been used for decades to predict outcomes probabilistically.
Monte Carlo Tree Search is easy to implement in code and is a good alternative to more traditional approaches like Minimax Tree with Alpha-Beta pruning. It uses probabilistic methods to find the seemingly best move, instead of heuristic analysis like Minimax.
According to Wikipedia: In computer science, Monte Carlo tree search (MCTS) is a heuristic search algorithm for some kinds of decision processes, most notably those employed in game play.
Programming sites like HackerEarth and Codingame hold Artificial Intelligence contests through games often. These are avenues for us to try the Monte Carlo Tree Search.
Mistake at 12:07 : The exploration bit is the second term. The exploitation bit is the value V_i.
Code:
https://github.com/gkcs/Competitive-P...
Links for Monte Carlo:
https://jeffbradberry.com/posts/2015/...
https://deepmind.com/research/alphago/
https://en.wikipedia.org/wiki/Monte_C...
http://mcts.ai/about/index.html
http://beej.us/blog/data/monte-carlo-...
Link for Minimax and Alpha-Beta:
Minimax: What is the Minimax Algorithm? - Arti...
Alpha-Beta: What is the Alpha Beta algorithm? - A...
Some Good chess Games where this might fail:
http://www.chessgames.com/perl/chessg...
http://www.chessgames.com/perl/chessg...
Strawpoll for future topics:
https://strawpoll.com/ycgc6ka7
P.S: The editing has been pretty rough on this video. Hope it doesn't impede the learning :)
Monte Carlo Tree Search is easy to implement in code and is a good alternative to more traditional approaches like Minimax Tree with Alpha-Beta pruning. It uses probabilistic methods to find the seemingly best move, instead of heuristic analysis like Minimax.
According to Wikipedia: In computer science, Monte Carlo tree search (MCTS) is a heuristic search algorithm for some kinds of decision processes, most notably those employed in game play.
Programming sites like HackerEarth and Codingame hold Artificial Intelligence contests through games often. These are avenues for us to try the Monte Carlo Tree Search.
Mistake at 12:07 : The exploration bit is the second term. The exploitation bit is the value V_i.
Code:
https://github.com/gkcs/Competitive-P...
Links for Monte Carlo:
https://jeffbradberry.com/posts/2015/...
https://deepmind.com/research/alphago/
https://en.wikipedia.org/wiki/Monte_C...
http://mcts.ai/about/index.html
http://beej.us/blog/data/monte-carlo-...
Link for Minimax and Alpha-Beta:
Minimax: What is the Minimax Algorithm? - Arti...
Alpha-Beta: What is the Alpha Beta algorithm? - A...
Some Good chess Games where this might fail:
http://www.chessgames.com/perl/chessg...
http://www.chessgames.com/perl/chessg...
Strawpoll for future topics:
https://strawpoll.com/ycgc6ka7
P.S: The editing has been pretty rough on this video. Hope it doesn't impede the learning :)
7 سال پیش
در تاریخ 1396/10/03 منتشر شده
است.
64,653
بـار بازدید شده